Every company supports a level of tribal knowledge. While some companies have more tribes than others, historically we know that tribes tend to bicker and skirmish on the borders of their domains. Best in class companies reward the pioneers who not only trap problems, but teach others to do so. It’s the end of the month and you are behind on shipments. An unexpected test result occurs and you can hear the groan from manufacturing a mile away. They know they did everything right. All materials passed QC inspection and the process was run without a deviation, process alert or alarm. All the samples were pulled at the right times, and everything looked spot on. What could have gone wrong?
As everyone is envisioning a complicated resampling and retesting plan, the quality group runs an unexpected result report for the past month for the given product. It turns out there is an open CAPA related to sample extraction and poor handling that introduces contamination. There are 10 unexpected results tied to this CAPA. The CAPA has not yet been implemented, and it turns out the line technician was not aware that leaving samples uncovered for several minutes prior to lab transport introduces contaminants. It turns out the sample in question was probably left open for an undetermined period of time prior to lab transport most likely causing the suspect lab result.
Due to this preliminary investigation, a much simpler retesting plan is quickly developed and approved. The test results come back as expected and a lengthy and expensive out-of-specification (OOS) investigation has been averted. The product is released to the warehouse without any out-of-tolerance (OOT) or nonconformance holds. We are much closer to our month end targets. Additionally, the unexpected result and quality event is tied to the pending CAPA, which avoids the creation of a separate CAPA.
Without a centralized quality management repository, previous unexpected lab results, process deviations, previous investigations and resulting CAPAs would have been stored in separate silos. In an attempt to meet month- end production commitments, an expensive retesting plan would have been developed potentially creating additional challenges, delays and overtime.
It is imperative to track process deviations, out-of-specification (OOS)/out-of-trend (OOT) results and product nonconformances. Unfortunately, the tracking of unexpected results and quality incidents tend to be relegated to Excel or small custom databases. These databases make it easy for end users to quickly enter data for later analysis. The data is separate from any Part 11 compliant QA tools and safe from any IT governance. These seemingly “quick and dirty” tools actually create larger, more subtle issues.
We have to admit there are constant tribal border skirmishes between QA, IT and Operations. Separate tribal knowledge and processes result in organizational inefficiency, extended product release times, missed production schedules, and unscheduled overtime. Without a centralized EQMS, collaborative pioneering degenerates into tribal data skirmishes (i.e. my data is better than your data). There will always be organizational conflict; however, a centralized EQMS approach most effectively shifts the debate from the source of the issue to what we are going to do about it.
Connect with Mike on Google+.